NTSB Glass Cockpit Safety Study Concludes Pilots Need More Training

 On Tuesday March 9, the NTSB held a public Board meeting to consider a study on whether glass cockpit-equipped general aviation aircraft are safer than aircraft with traditional round gauges. The Board reached the same conclusion as other similar studies. Glass Cockpits have the potential to increase safety. However pilots need additional training, both initially and on a recurrent basis, to reap the benefits of these technologies. Five of the six study recommendations were...

 related to training. 

As a flight instructor who specializes in teaching in glass cockpits I’m not surprised. While glass cockpits have dramatically increased the capabilities available to pilots, they have also increased the amount of training required to be proficient with those systems. 

And that’s a conundrum for General Aviation. At the same time we’re trying to attract more people to aviation, the learning curve for becoming proficient with glass cockpits is getting steeper. Ultimately, I expect user interfaces will improve and workloads will decrease. But until then, pilots flying glass cockpits must invest the time and money to seek out the best possible training (see end of article for more on this). Not only do pilots need good initial training on using the systems and understanding failure modalities, but the Board feels they also need better recurrent training.

Study Background

The goals of the study were simple. The NTSB wanted to identify any difference in operational characteristics of glass and non-glass aircraft and determine how glass has affected safety. To do this, they decided to compare accident information and activity between matched groups of aircraft, specifically glass and non-glass aircraft manufactured between 2002 and 2006. The groups were selected to minimize any differences in which the two groups of airplanes used. In the end however, the study found that there are very large differences in the way glass and non-glass aircraft are used.

The study group included the Cessna 172, 182, 206, 350 and 400, the Cirrus SR20 and SR22, the Diamond DA40, Mooney, Beechcraft A36 and G36 and numerous Piper models. In the five-year period, a total of 5516 glass and 2848 non-glass aircraft of these models were manufactured. The NTSB looked at the accidents these aircraft had between 2002 and 2008.

Study Results

The study aircraft had a total of 266 accidents. In glass aircraft, there were a total of 125 accidents of which 39 were fatal. In non-glass aircraft, there were a total of 141 accidents of which 23 were fatal. 

Simply counting the number of accidents that occur is misleading, since it doesn’t account for variations in the number of hours that different types of aircraft fly. To compute accident rates, the NTSB worked with manufacturers to determine the average number of hours flown per year by model. Using this data, they calculated a glass cockpit total accident rate of 3.77 accidents per 100,000 hours versus 3.71 for non-glass aircraft. 

Surprisingly the NTSB didn’t compare these accident rates with those computed for the entire fixed wing general aviation fleet. AOPA’s Air Safety Foundation supplies those figures in their annual Nall Report, an authoritative source of general aviation accident data (commercial operations conducted under FAR Part 135 are not included in the Nall Report). Nall Report data for 2007 computes a total accident rate of 6.47 per 100,000 hours, nearly twice the rate in the NTSB study. The simple conclusion is that new aircraft manufactured between 2002 and 2006 are twice as safe as the entire GA fleet, which includes some aircraft that are more than 80 years old.

For fatal accidents, the story was quite different. The study found the glass cockpit fatal accident rate was 1.03 accidents per 100,000 hours versus 0.43 for non-glass aircraft). Nall Report data for 2007 computes a fatal accident rate of 1.18 per 100,000 hours.

Further Analysis

It would be easy to leap to the conclusion that glass cockpit aircraft are less safe except for one fact: glass cockpit aircraft are used very differently than their non-glass counterparts, exposing them to greater risk. The Board found that glass cockpit aircraft were more likely to be used for business and personal flying and to be flown IFR by older pilots. Non-glass airplanes had a significantly higher percentage of takeoff and landing accidents, were more likely to be used for instruction and were flown on shorter flights. Glass aircraft averaged trips of 96 miles, exposing them to more weather and terrain dangers, while non-glass aircraft averaged trips of just 25 miles.

Again, the Board missed an opportunity to compare their study data with Nall Report data for different types of flying activity. Instructional flying is relatively safe, since it’s done under controlled conditions with a flight instructor supervising the flight. Nall Report data for 2007 show that instructional flights accounted for 15.2% of hours flown, but only 5.9% of fatal accidents. Applying this ratio to overall fleet data suggests an instructional fatal accident rate of 0.458 accidents per 100,000 hours, nearly identical to the rate the Board found for non-glass aircraft, which were doing a high percentage of instructional flight. Combined 2007 Nall report personal and business flight data accounted for 53.5% of hours flown, but a whopping 77.2% of fatal accidents. Applying this ratio to overall GA fleet data suggests an overall personal and business fatal accident rate of 1.70 accidents per 100,000 hours. This implies that glass cockpit aircraft, when matched against aircraft used for a similar purpose, are actually safer than the overall GA fleet, which is predominantly non-glass.

NTSB Board member Robert Sumwalt zeroed in on this issue when he asked “This is the million dollar question. Were the differences because of the aircraft or the mission?” A staff member replied that he “couldn’t say for sure.”

 

Recommendations

The NTSB Board came up with six recommendations, five of which were training related. They were that the FAA:

1. enhance pilot knowledge and training requirements;

2. require manufacturers to provide pilots with information to better manage system failures;

3. incorporate training elements regarding electronic primary flight displays into training materials and aeronautical knowledge requirements;

4. incorporate training elements regarding electronic primary flight displays into initial and recurrent flight proficiency requirements for pilots of small light general aviation airplanes equipped with those systems, that address variations in equipment design and operations of such displays;

5. support equipment-specific pilot training programs by developing guidance for the use of glass cockpit simulators other than those that are approved by the FAA as flight training devices; and

6. inform the general aviation community about the importance of reporting malfunctions or defects with electronic flight, navigation and control systems through the Service Difficulty Reporting system.

Your Personal Training Program

 If you’re flying or planning to fly glass cockpit aircraft, you’ll want to get the best training available. If I can be of help to you in tailoring a program to fit your needs, please contact me at 650-224-7124. I spend 100% of my time teaching in glass cockpit aircraft and have access to three G1000 simulators, including the latest Redbird motion simulator. I teach at the Palo Alto airport, near San Francisco, the #1 travel destination in the world. Or, I’ll be happy to fly with you at your home airport. Regardless, give thought as to how you plan to improve your flying proficiency in 2010. Fly safely!

Comments

[Mike Danko]

Max- Aircraft with glass cockpits may indeed be used more for business travel than would similar aircraft with legacy instrumentation. But that doesn't mean the glass cockpit safety record shouldn't be compared to that of the legacy aircraft population as a whole. A glass cockpit "invites" a pilot to attempt business trips in adverse conditions -- conditions that perhaps would have resulted in a cancellation if the particular pilot had only a legacy aircraft at his disposal. That factor needs to be accounted for, and we need to be careful not to wash it out of the analysis.

More here => http://www.aviationlawmonitor.com/2010/03/articles/ntsb/ntsb-glass-cockpits-associated-with-higher-rate-of-fatal-accidents/

[Mstibbe]

It would be interesting to compare the accident rates for things where glass cockpits add functionality (e.g. instrument approaches, weather avoidance, CFIT and midair collisions) and things where they don't (e.g. mechanical problems, mishandling, fuel mismangement etc.)

My suspicion is that we get false comfort from glass cockpits because they protect us from *some* risks but, in fact, they don't protect us from the risks that cause the most accidents.

In other words, while extra training on glass cockpits would be a Good Thing, it needs to be matched with good basic handling, airmanship, fuel management and so on.

I've know Cirrus pilots who almost never 'fly' the plane. They rely on the autopilot and avionics to do the work. This is fine if they have good flying skills in the background but, quite frankly, they don't. A couple of people I know probably couldn't recover safely from a stall or do a steep turn on instruments. They think I'm mad because I like to hand fly the plane most of the time.

The risk is that if you fly once or twice a month in a very sophisticated plane and all you do is go from one big airport to another with IFR flight plans, everything looks great and you feel like a shit-hot pilot until something goes wrong and your lack of real currency catches up with you.

I've had alternator failures, mag failures, PFD and MFD failures, unserviceable autopilot failures etc. It really does happen. I had to hand-fly into Le Bourget single pilot IFR when my autopilot failed. My big fear is that my ability to do this and keep my skills current is fading over time.

I suppose another concern is with differences between glass cockpit systems. I'm very comfortable with Avidyne + Garmin 430s and some of those skills are transferable to the G1000 but it still took me four hours or so to get a basic level of familiarity with it when I flew with you in January. I suspect 40-50 hours plus lots of ground study is required for complete fluency. But then they go and fit Avidyne R9 systems and you have to start again. Hmmm.